Templates

There is a lot to cover here but for now we will cover the basics.

Branches most fundamentally define Templates - either an Agent Template or a Scenario Template. These Templates then define reusable, instantiatable definitions for an Agent or Scenario. To recap, an Agent is an actor within a simulation and a Scenario defines the simulation (including the Agents that make it up and the Templates that define them). Therefore, Agent Templates capture the model for a system while Scenario Templates capture the instantiation of differentiated Agents, the resolution of their abstract "Targets", and the configuration of the simulation clock.

A Template is made up of Blocks - each existing within a Block Group inside of the Template. A Block Group is much like a traditional database table but with more flexibility. The following is the default Scenario Template:

{
    'ClockConfig': {}, # Block Group
    'Agent': {}, # Block Group
    'Orbit': {}, # Block Group
    'clockConfig': None, # root property (in this case a relation to a ClockConfig Block that defines the start and stop time of the simulation)
}

Templates are modified by creating, updating, and deleted their Blocks using the "Blocks" endpoints below.

Template Composition

Scenario Template

• Agent
• ClockConfig
• Orbit

Agent Template

All Blocks not referenced above under "Scenario Template". Fetch one of the Agent Template branches that was pre-loaded into your account for a complete example.

Queries

Data is queried via 4 URL query parameters:

• id: The ID of the data set to query. This ID is defined on a Simulation Job.
• start: The start time of the window in MJD. If omitted, defaults to the minimum timestamp in the dataset.
• stop: The stop time of the window in MJD. If omitted, defaults to the maximum timestamp in the dataset.
• binWidth: (optional) The bin width to use to down-sample the data. If omitted, data will be full resolution.

Example request: GET: /data/?id=<id>&start=<start>&stop=<stop>&binWidth=<binWidth>

Response

Data is returned from a query in the following format:

{
    "series": { 
        <streamId>: [
            [<timestamp1::MJD>, <timestamp2::MJD>, ...],
            {
                <agentId>: {
                    <blockId>: {
                        <stateVariableA>: [<valueA1>, <valueA2>, ...],
                        <stateVariableB>: [<valueB1>, <valueB2>, ...],
                        ...
                    },
                    <stateVariableC>: [<valueC1>, <valueC2>, ...]
                },
                ...
            }
        ],
        ...
    }
}

Note that a "Stream" is a set of state variables that share the same timestamps. Specifically, the values in a Stream correspond to a modsim Engine of a particular Agent. The series object contains a key for each stream in the data set. The value of each stream key is a 2-element array. The first element is an array of timestamps - the X-axis for all variables in the stream. The second element is an object containing the state variables that compose the stream. The state variables are grouped under an Agent ID and then a Block ID (if applicable). The value of each entry is an array of values that aligns with the timestamp vector.